Browsing School of Electronic Engineering and Computer Science by Author "Ummadisingu, A"
Now showing items 1-2 of 2
-
Hindsight policy gradients
Rauber, P; Ummadisingu, A; Mutz, F; Schmidhuber, J (2019) -
Reinforcement Learning in Sparse-Reward Environments with Hindsight Policy Gradients
Rauber, P; Ummadisingu, A; Mutz, F; Schmidhuber, J (Massachusetts Institute of Technology Press (MIT Press), 2021-05)A reinforcement learning agent that needs to pursue different goals across episodes requires a goal-conditional policy. In addition to their potential to generalize desirable behavior to unseen goals, such policies may ...